58 Transportation Research Record 1043

Investigation of Seasonal Load Restrictions 1n Washington State

JOE P. MAHONEY, JO A. LARY, JAY SHARMA, and NEWTON JACKSON

ABSTRACT

Presented in the paper are the results of monitoring seasonal changes in surface deflections, temperatures, moisture contents, and calculated layer moduli for six test sites in the state of Washington. The goal of the data collection and analy­sis was to evaluate the Washington State Department of Transportation (WSDOT) load restriction tables used on some state routes during the spring thaw. Further, a criterion was developed to estimate when seasonal load restrictions need to be applied to those pavement sections that require them. Extensive use was made in the study of the WSDOT falling weight deflectometer to obtain pavement surface deflection basins and of multilayered elastic computer programs to analyze the data.

Pavement engineers throughout much of the United States are faced with the recurring problem of weak­ened pavement structures during the spring thaw. To reduce the pavement deterioration that can occur during this period, load restrictions for truck traffic are often applied. A survey reported in NCHRP Synthesis Report 26 (!.1 p.77) revealed that slightly less than one-half of the states use sea­sonal load restrictions. When such restrictions are used, several questions arise and can include the following:

1. Which pavement sections require load restric­tions?

2. When should restrictions be applied and re­moved?

3. For Washington State, are the present load restrictions (developed in 1952) adequate, and how do they affect groups such as freight and timber hauling companies and school buses?

The current load restriction tables used in Wash­ington State are summarized in Table l and can be applied at two levels:

1. Emergency load restrictions and 2. Severe load restrictions.

TABLE 1 Current WSDOT Load Restriction Tables•

Conventional Tires Tubeless Tires

Allowable Gross Load (lb) Allowaule Gwss Luad (lb) Tire Width Severe (in. ) Emergency Emergency

7.00 1,800 1,800 7.50 2,250 1,800 8.25 2,800 1,900 9.00 3,400 2,250

10.00 4,000 2,750 11.oob 4,500 3,000 12.00C 4,500

aLast revjsed October 1957. bor more-severe emergency condition. CQr more-emergency condition.

Tire Width Severe (in.) Emergency Emergency

8-22.5 2,250 1,800 9-22.5 2,800 1,900

10-22.5 3,400 2,250 11-22.5 4,000 2,750 11-24.5 4,000 2,750 12-22.5b 4,500 3,000 I 2-24.5c 4,500

To date, the Washington State Department of Trans­portation (WSDOT) has applied such load restrictions primarily on the basis of experience and occasionally on the basis of either Benkelman beam or falling weight deflectometer (FWD) pavement surface deflec­tions. Most load restrictions are applied to low­traffic-volume routes such as the Federal Aid secon­dary system. Further, most counties use identical load restrictions and application periods.

The overall objective of the reported research was to evaluate the effect of freeze-thaw in pavement layers on pavement structural capacity. More spe­cifically the objectives were to

1. Measure the variation of base and subgrade moisture content, frost depth and location, and pavement deflection (surface and in situ):

2. Develop procedures for using easily obtained data or otherwise provide for predicting when load restrictions should be applied on a given pavement structure: and

3. Determine an appropriate load restriction criterion.

To accomplish these objectives it was necessary to

1. Collect data at several test sites, including measurement of

• Frost depth using frost tubes, • Moisture contents using soil cells, • Soil temperature using soil cells, • Dynamic deflection basins using the FWD, • Static deflections using a Benkelman beam, and

Dynamic and static defleclions using an exten­someter permanently buried in the pavement structure.

2. Collect weather data. These data, obtained from National Oceanic and Atmospheric Administration climatic reports or the WSDOT maintenance offices, were used to calculate freezing indices and to esti­mate depth of freeze using the modified Berggren equation.

3. Obtain pavement samples. Samples of the base and subgrade materials and cores of the asphalt con­crete were obtained for laboratory resilient modulus determination. At the time of sampling, the in situ

Mahoney et al.

FIGURE l Location of field test sites.

density and moisture content of the base and sub­grade were determined.

FIELD STUDY

Site Se l ecti o n

Test sites on existing WSDOT routes were chosen in the central part of Washington (WSDOT District 2). In this part of the state the design freezing index i s about 1,000 degree-days (Fahrenheit). Several criteria were used as a basis for selecting the test sites and include

l. The pavement must be located in an area of potential deep frost and be plowed in the winter to keep it free of snowi

2. Each test site should have a reasonable amount of heavy truck traffic (weight but not necessarily volume) i and

TABLE 2 Principal Test Site Features

State Route and Mile Post Instrumentation Pavement Structure•

SR 97, MP 2 frost tubes 4-in. ACP 184 I moisture tubeb 4-in. CSTC

6-in. ballast SR 2, MP 2 frost tubes 6-in. ACP

117 2 moisture tubes 17-in. gravel base 1 extensometer

SR 2, MP 2 frost tubes 2-in. ACP 160 1 moisture tube 9-in. CSTC

SR 172, MP FWD testing 2.6-in. BST 2 6-in. gravel base

SR 172, MP FWD testing 2-in. BST 21 9-in . gravel base

SR174,MP 2 frost tubes 0.5-in. BST 2 1 moisture tube 2-in. ACP

1 extensometer 10-in. gravel base

59

Spokane

~ =TestSite

3. The site locations should encompass a variety of subgrade soil and drainage conditions.

Using these criteria as a partial basis, six 500-ft­long test sites were selected (locations shown in Figure l) for deflection testing, and four of the sites were instrumented. An overview of the signifi­cant test site features is given in Table 2.

Data Collected

Field data were collected at the six test sites dur­ing a 15-month period beginning in January 1982, with special emphasis on the spring thaw period. The following data were collected:

1. Pavement surface deflection using the FWD or Benkelman beam, or bothi

2. Extensometer readings;

Sub grade Traffic, 1982 (2)

Percentage Passing Percentage

Class 200 Sieve ADT Trucks

A-1-a(O) 9 3,500 11

A-1-b(O) 16 to 19 11 ,500 10

A-1-a(O) 9 to 12 1,000

180 6

530 6

A-1-b(O) 18 to 22 820 16

8Nomencl~ tura: ACP = asphalt con ere te pavomont, CSTC =crushed surfadng top course, and BST = bi luminous surface tree tmenl. b£ach mois tLJte tube consisted of four moi:11urc cells.

60

Asphalt Concrete

Transportation Research Record 1043

Flexible tube lor retrieval and to allow for expansion

RemovabiB inmH PVC tube, 7/8" 111

Permanently installed outer polyethylene tube, 1-1/4" 111

Ottawa sand saturaled wilh 0.1 % solution fluorescein dye

FIGURE 2 Schematic of in situ frost tube.

3. Pavement surface temperature; 4. Base and subgrade temperature; 5. Soil cell resistivity (for moisture content

determination); and 6. Depth of frost penetration (using frost

tubes).

Instrumentation

A primary objective of the study was to measure changes in pavement strength over the project dura­tion. To this end, four of the six test sites were instrumented (as indicated in Table 2) with frost tubes to measure depth of freezin<J and s o il cells (Soiltest MC 310A) to measure subgrade and base course moisture contents and temperatures. Exten­someters we re installed at two sites to measure pavement surface deflections. Paint marks were placed on the pavement surface to facilitate repeating de­flection testing at the same locations. Sketches of

..--- ------- Lead wire to Bison measuring instrument

3" 0 aluminum plate

1" 0 Bison coils

1" i.d. PVC pipe

Note: Top extensometer plate embedded in cold mix, 3-1/2" from pavement surlace

1/2" 0 sleel rod

3" 0 aluminum plate

FIGURE 3 Schematic drawing of extensometer.

typical frost tube and extensometer construction and soil cell layout are shown as Figures 2-4.

Deflection Measurements

With the advent of computer programs that provide for using pavement deflections to estimate in situ pavement layer moduli, the usefulness of devices that provide such measurements continues to increase. Three approaches were used in the study to measure pavement deflections:

1. Benkelman beam, 2 . FWD, and 3. Extensometer.

Soil r.ells

FIGURE 4 Typical soil cell layout .

Removable cap

1-114" polyethylene tube grouted in place

Mahoney et al.

The Benkelman beam was used to measure single-point rebound deflections along with a single axle, dual tire 18,000-lb axle load (quasi-static loading). The FWD was used to measure deflection basins (with seven sensors). The FWD can apply a dynamic load ranging from 3,000 to 24,000 lb and simulates a vehicle mov­ing at speeds greater than 30 mph. Finally, the ex­tensometer was used to measure deflections under various axle and tire loadings. The Benkelman beam and FWD measurements were taken at 50-ft intervals at each test site.

Typica l Results

The type and range of the data collected from the test sites will be illustrated by use of two of the six test sites [SR 2, milepost (MP) 160 and SR 172, MP 2].

12 35

11 Deflection, normalized to 9,000 lb load

10 30

9

8 25

~ :£ "? ~

~ 20 .2

Subgrade moisture content u _gi at 4 ft . depth Q; 0

~ 15

c: 7 " c: 0 () 6 ~ :0

.11i 0

::::;

4

3 Basa moisture content at 1 ft . depth

10

Subgrade moisture content at 2 ft . depth

o 5 F M A M

1983

61

FWD Deflections

For the two selected test sites, the maximum pavement deflection (first sensor) averaged over each test site and normalized to a 9,000-lb load is plotted versus time in Figures 5 (SR 2, MP 160) and 6 (SR 172, MP 2). According to these data, spring was the period of highest deflection, as expected (similar trends were observed for the four other test sites). In general, maximum deflections were reached in late February or early March. During the January 1984 site visits, the measured deflections ranged from 9 to 20 percent of the previous summer values, which illus­trates the increased stiffness of frozen pavement layers .

Figures 5 and 6 were several site visits and trends (as plotted) mask occurred at other times available.

A s 0 N D F

developed on the basis of corresponding data. These the actual variations that

for which data are not

M 1984

Time

FIGURE 5 SR 2, MP 160-plot of FWD first sensor deflection and base and subgrade moisture contents versus time.

80

60 g

"? Deflection, normalized ~ to 9, 000 lb load ~ c: 40 .ll

~ Q; 0

20

0 F M A M J A S 0 N D F M

1983 1984

Time

FIGURE 6 SR 172, MP 2-plot of FWD first sensor deflection versus time.

62

Base and Subgrade Moisture Contents

It is generally accepted that soils (and asphalt concrete) exhibit a decrease in strength with in­creasing moisture content. To this end, soil resis-tivity ~a~ mca~ured and later converted to soil moisture content when the appropriate laboratory calibrations were completed. An example of the annual change in soil moisture content is shown in Figure 5. Moioture content tends to increase with depth (a» expected) and is most variable during the thawing period (winter-spring 1984). Of significance is the high moisture content of the base course, which was essentially as high as the moisture content in the subgrade located 2 ft below the measurement point in the base, This will be further discussed later.

Temperature

As previously discussed, a variety of measurements was made to relate air and pavement temperatures. Soil celis were used to measure in situ soil tem­peratures as well as soil resistivity for moisture determination. Frost tubes were used in an attempt to measure locations of frost up to a depth of 4 ft; however, numerous problems arose with these and the resulting data were of little use.

On the basis of air temperature data from the nearest WSDOT maintenance facility, freezing indices were calculated and are given in Table 3. These data show the differences in just two winters (1982-1983 and 1983-1984). The design freezing indices (the average of the three coldest winters out of the last 30 years of record) range between 900 and 1,100 degree-days through the part of Washington State where the test sites are located (the mean freezing index for a 30-year record is about 500 degree-days) • Thus the winter of 1982-1983 was slightly less severe than average and the winter of 1983-1984 was above average.

By use of the modified Berggren equation (3), various estimates of the depth of ground freezing were made. Assuming the pavement structure and the upper portion of the subgrade can be characterized as a homogeneous granular material (ya = 130 pcf, w = 5 percent) for thermal purposes, depths of freeze were calculated for the appropriate freezing indices at each test site and are given in Table 3. Also given in this table are the measured depths of frost (soil cells). The calculated and measured values are at least within the same range (recall that the soil cells were placed at 1-ft intervals of depth, making such comparisons quite approximate). Overall, the temperature data and associated calculations suggest that typical depths of freeze beneath the test sites

Transportation Research Record 1043

are about 3 to 4 ft with maximum frost penetration occurring in late January through mid-February. As will be shown in the next section, relatively few thawing degree days are required to place these pavement structures in a "critical period."

DATA ANALYSIS

The method for determining pAvPmPnt matPrial proper­ties and their variation with time for the six test sites is described in this section. Further, devel­opment of a temper a ture-based er i ter ion for es tab-1 ish ing when to apply load restrictions is presented along with the procedure for determining the magni­tude of restrictions.

Material Properties

The BISDEF computer program !il was used along with FWD data to estimate the moduli (or resilient moduli, MR) and the associated stress sensitivity relations for the pavement layers. This program is based on layered elastic theory and was developed at the U.S. Army Corps of Engineers Waterways Experiment Station. It uses the concept of minimizing the difference be­tween the program-calculated and the measured de­flection basins. The program varies the layer modulus until a match is made between the input basin and the BISDEF-predicted basin within a specific margin of error.

The modulus for each pavement layer was estimated for each site visit and test site. The program was also used to calculate bulk stress at the middle of the base course and bulk or deviator stress at the top of the subgrade. Because a minimum of three stress levels (or FWD load magnitudes) was used during each site visit, it was possible to develop the stress-sensitivity relationships (MR - 8 or MR -aa) for the base and subgrade layers. These relation­ships were necessary for additional modeling, which will be discussed later.

The required inputs for the BISDEF program were

1. Measured deflections (mils) and associated distances from the center of the load (inches).

2. Range of modulus values for each layer (psi), 3. Initial estimate of the modulus value for each

layer (psi), 4. Thickness of each layer (inches), 5. Poisson's ratio for each layer, 6. Load stress (psi) and load radius (inches),

and 7. Points in the pavement structure where

stresses are desired.

TABLE 3 Freezing Indices and Calculated Depth of Frost Penetration for the Six Test Sites

Freezing Index (degree-days

Test Site Winter Fahrenheit)

SR 97, MP 184 1982-1983 475 1983-1984 685

SR 2, MP 1 17 1982-1983 1983-1984 510

SR 2, MP 160 1982-1983 400 IY83-1 984 '/30

SR 172, MP 2 1982-1983 475 1983-1984 745

SR 172, MP 21 1982-1983 475 1983-1984 745

SR 174, MP 2 1982-1983 170 1983-1984 470

asased on recorded temperature data.

Calculated Depth of Frost (ft)

3.3 4,0

3.4 3.1 4.1 3.3 4.2 3,3 4.2 2.0 3.3

Minimum Measured Depth of Frost (ft)

s 3+

3+

2+

Prolrnhle Date of Maximum Depth of Frost'

Feb. 9, 1983 Jan. 23 , 1984

Jan. 23, 1984 Feb. 11, 1983 Jan. 23, 1984 Feb. 15, 1983 Jan , 23, 1984 Feb. 15, 1983 Jan. 23, 1984 Jan. 5, 1983 Jan. 24, 1984

Mahoney et al.

The measured deflections input to the program were the average deflection basins over each test section for each site visit. The input deflections were selected from four of the FWD sensors and were lo­cated at spacings of O, 11.8, 25.6, and 47.2 in. from the center of loading. An initial estimate of each layer modulus was based on judgment and other previously completed work. The load radius was that of the FWD loading plate (5. 9 in.) and the load stresses were the actual stresses applied to the pavement structure by the FWD.

The results of the BISDEF analysis are given in Tables 4 and 5 for the two test sites used in this paper to illustrate the study findings (SR 2, MP 160 in Table 4 and SR 172, MP 2 in Table 5). The stress

relationships presented are of the form MR = k1ek2

for layers that behaved as coarse-grained materials -k2 and MR = k1crd for layers that behaved as fine-

grained materials (MR decreasing with increasing stress.) The SR 172, MP 2 test site was run as a two-layer system with the bituminous surface treat­ment and base course combined into one layer. This was done because the computer program would not close when the site was run as a three-layer system.

In general, for all test sites as well as the two sites shown, the base and subgrade moduli were higher for the August 1983 site visit than at other times of the year (as might be expected). Further, the pavement layer moduli were substantially higher when frozen. An interesting trend for SR 2, MP 160 was that the base modulus decreased 41 percent from August 1983 to March 1984, but the subgrade modulus on both dates was about the same. For SR 172, MP 2,

63

the base modulus decreased about 27 percent and the subgrade modulus 44 percent f or the same time inter­val (this test site had the largest decrease in sub­grade modulus of the six). The maximum observed decrease i n modulu s for the base· was 78 percent (August 1983 and March 1984 testing dates). For all test sites (except SR 2 , MP 117, whi ch exhibited ex­tensive fatigue crack i ng and was actually weaker during the summer month s ), the base course modulus was reduced by an average of about 52 percent and the s ubgrade modulus by about 23 percent.

Magnitude of Load Restrictions

The PSAD2A computer program (5) was used to calculate deflections and strains under-a given wheel load for the sununer (strongest condition) and the spring (weakest condition) for each of the test sites. This was done to determine the change in strains and de­flections between the two cases so that a spring load could be found that induced the same strains and de­flections, and hence potential pavement damage, as occurred in the sununer under maximum loading.

Several input values were required for the PSAD2A program and included for each layer

l . Poisson's ratio, 2. Dry density, 3. Thickness, 4. Stress-modulus relationship (from BIS DEF

analysis) , and 5. Initial estimate of modulus for each layer.

Because the vast majority of trucks uses tubeless tires and the maximum wheel load is in part a func-

TABLE 4 SR 2, MP 160-Results of BISDEF Analysis for Determination of Resilient Moduli, Stresses, and Stress Relationships for each Site Visit

Applied Subgrade Subgrade Tempera- Stress AC MR Base MR MR Base 0 Oct Base Stress Subgrade Stress

Date ture (°F) (psi) (psi) (psi) (psi) (psi) (psi) Relationship Relationship

02/24/83 50 57.86 1,200,000 14,800 13,900 21.86 11.67 1 1,26600,814 29,862 Oct -0.307 88.17 1,500,000 20,800 13,100 28.42 15.85

r2 = 0.962 r2 = 0.944 123.01 1,200,000 29,400 11,700 49. 18 20.29 03/03/83 45 56.72 l ,200,000 21,600 13,100 18.94 10.51 J 6,70600.396 17,462 Oct-O.J ! 8

83.06 l ,200,000 24,400 13,000 26 .53 14.96 121.37 1,268,000 27,500 12,100 34.75 20.38 r2 = 0.997 r2 = 0.791

03/09 /83 47 55.80 l,000,000 25,300 13 ,500 18.98 10.51 l 9,50880.327 25,782 Oct - 0·268 82.44 1,222,000 26,400 13,000 25.7 1 14.62

116.70 1,300,000 30,000 11,500 31.02 18.74 r2 = 0.836 r2 = 0.864

03/13/83 60 42.70 I, 100,000 15,200 10,000 13 .90 7.27 { 2,78700.65 8 21,904 Oct -O. I 9 9

83.96 1,029,000 26,800 12,700 26.90 15.10 121.83 1,267,000 28,000 12,200 35.08 20.53

r2 = 0.953 r2 = 0.926 152.87 1,318,000 30,500 11,500 40.01 24.33 03/24/83 40 58,52 2,155,000 16,800 13,000 17.84 9.90

{ 4,24580.4 74 21,088 Oct-0. 20l 88.36 2,460 ,000 18,800 12,800 25.05 14.20 125.46 1,600,000 26,300 12,200 34.44 20.41

r2 = 0.669 r2 = 0.632 150.67 2,400,000 20,800 10,600 36.41 22.05 08/17/83 72 80.08 931,000 28,800 11,100 23.08 13.50 } 282&' ·473 6,152 Oct0 ,227

125.98 1,000,000 43,300 12,000 30.44 19.04

01/10/84 34 76.63 108.83 144.39

02/21/84 42 71.0 1,096,000 20,800 12,900 24 .54 13 .50 ~ 6,27780. 376 21,492 od- 0·197 95.7 1,140,000 23,000 12,200 30.73 17.30

r2 = 0.988 r2 = 0.994 129.2 1,258,000 24,600 11,700 38.35 22.16 03/01/84 48 71.3 1,184,000 16,500 12,500 25.30 13.75 1 4,62800, 39 8 26,576 Oct-0· 287

95.9 1,327,000 18,800 11,800 31.05 17.31 125.8 1,462,000 19,300 11,000 37.60 21.46 r2 = 0.886 r2 = 0.994

03/09/84 60 66.6 972,000 24,700 12,800 22.26 12 ,39 J 4,4618°·554 26,189 Oct-0.282

91.0 552,000 30,400 12,000 30.80 16.83 121.8 579,000 32, 700 11,000 37.29 21.03 r 2 = 0.984 r2 = 0.968

03/21/84 49 66.0 658 ,000 30,300 12,900 22.53 12.36 l 18,5040°· 1 5 J 17,637 Oct-0 .! 22

93.3 819,000 29,600 12,600 30.00 16.86 127.1 700,000 33,000 12,000 39.20 22.19 r2 = 0.536 r2 = 0,947

64 Transportation Research Record 1043

TABLES SR 172, MP 2-Results of BISDEF Analysis for Determination of Resilient Moduli, Stresses, and Stress Relationships for each Site Visit

Applied Temp~ra- Stress !st Layer Subgrade

Dat~ t11rP- ( F) (psi) Mp_ (psi) MR (psi)

02/24/83 50 47.46 13,400 5,000 72.6 9 14,800 5,700

103 .32 24,300 5,700

03/03/83 38 49.54 25,900 5,800 71.65 28,000 5,900

105 .24 29,900 5,900

03/09/83 47 47.95 28,800 6,800 68.40 31,000 6,700

101.05 33,500 6,600

03/17/83 39 49.22 21,200 6,600 70.6 8 23,700 6,600

102.86 22,800 6,700

08/17/83 75 74.70 26,000 8,600 105 .82 29,000 8,600

01/10/84 34 72.84 371,100 59,900 104.21 348,300 59,300 143.49 326,800 50,500

03/01/84 46 56.8 23,700 4,500 78.5 32,000 4,800

110.0 39,200 5,100

03/07 /84 60 55.2 19,900 5 ,300 74.9 19,000 5,500

103.9 27,400 5,800

03/21 /84 50 57.1 28,400 6,000 77.6 27 ,000 5,800

107.0 31,600 5,700

tion of tire width, it was decided that the following tire sizes would be used in the subsequent analysis: 8-22.5, 9-22.5, 10-22.5, 11-22.5, 12-24.5, 14-17.5, and 16-22.5. Only single tires on single axles were evaluated because these were considered to be the r.ios t er i ti cal.

For the sununer load cases, the maximum allowable load per time would be input. This maximum was determined by use of the Revised Code of Washington (RCW) 46.44.042, which allows 550 lb per inch width of tire up to a tire width of 12 in. and 660 lb per inch width for tires 12 in. wide or wider. For ex­ample, an 11-in.-wide tire can legally carry 6,050 lb and a 12-in.-wide tire 7 ,920 lb. Corresponding tire pressures were based on tire inflation pressures reconunended by the Tire and Rim Association, Inc.

TABLE 6 Tire Loads and Tire Pressures for the Spring Condition

Percen tagc of Maximum Load

100

75

50

Tfrc Size

8-22.5 9-22.5

10-22.5 11-22.5 12-24.5 14-17.5 16-22.5 8-22.5 9-22.5

10-22.5 l l-22.5 12-24.5 14-17.5 16-22.5 8-22.5 9-22.5

10-22.5 11-22.5 12-24.5 14-17.5 16-22.5

Tire Pressure Load/Tire (psi) (lb)

105 4,400 115 4,950 105 5,500 JOO 6,050 115 7,920 100 9,240 90 10,000 80 3,300 75 3,712 70 4,J :!) 65 4,538 80 5,940

100 6,930 75 7,500 55 2,200 55 2,475 55 2,750 65 3,025 65 3,960 65 4,620 55 5 ,000

!st Layer Subgrade I st Layer Stress Subgrade Stress 0 (psi) ad (psi) Relationship Relationship

47.58 15.91 l 61300.?81 2,593 ad o.2 39 73.09 24.62 97.88 28.70 r2 = 0.788 r2 = 0.937

4G.72 13.55 j 12,21:100.195 o,45 7 ad 0·024

67.02 19 .03 97.54 27.07 r2 = 0.996 r 2 =0 ,74 1

45.45 13.34 1 12,97800,209 7,634 ad - 0.045 64.20 18.40 93.67 26.01 r 2 = 0.999

48.25 15.34 1 14,99500.096 6,205 a d 0,02 I 68.36 21.06

100.23 31.43 r 2 = 0.386 r 2 = 0.806

73 .76 23.85 } 6,39300.326 8,600 ad 0·0

103.09 32.32

65.82 17.08 94.82 25 .06

128 .97 32.96

52.39 14.36 ! 9528°· 8 J? 2,268 ado.25s 70.38 17 .86 96.80 23.27 r2 = 0.982 r2 = 0.994

53.14 16.19 1 2,35500.520 3,352 ad0.163 72.79 22.69 97 .13 27.56 r2 =0.619 r2 = 0.935

53,42 15.20 13,7610°· 173 7,71 9 ad-0.093 72.80 20.82 98.14 26.49 r2 =0.432 r2 = 0.988

These pressures were found to be reasonable for modeling purposes on the basis of a previous study performed for WSOOT (&_). The tire loads and pressures for the sununer condition (maximum condition) are given in Table 6 (100 percent of maximum load).

For the spring condition, the following cases were developed:

1. The maximum load and tire pressure as used for the sununer condition,

2. Seventy-five percent of the maximum load and corresponding tire pressure as reconunended by the Tire and Rim Association, and

3. Fifty percent of the maximum load and the reconunended tire pressure.

The resulting tire loads and pressures are given in Table 6.

The output parameters from PSAD2A, which were evaluated for both the sununer and spring analyses, were

1. Surface deflection (o), 2. Horizontal strain at the bottom of the bi­

tuminous bound layer (Etl, 3. Vertical strain at the top of the base course

<'val, and 4. Vertical strain at the top of the subgrade

<'vs>

When these deflections and strains had been calcu­lated, the spring load that. r.aused the same damage as the maximum allowable load during the sununer was computed. This was done by use of plots developed from the previously listed program outputs for each test site and tire size and is shown for SR 2, MP 160 and tire size 11-22. 5 in Figure 7. This figure was constructed as follows:

1. Surface deflection versus load was plotted for the three load s used in the spring analysis and a curve was fitted through the points and

Mahoney et al. 65

Evs 25 1500

:? c: "' "E

'0

20 ·- 1000 '\'

" ~ c: Evs ~

~ c: '!!!

Qi li5 0 15 500

10

3 4 5 6 7

Tire Load (kips)

FIGURE 7 SR 2, MP 160-tire size 11-22.5.

2, £t1 £VB• and £VS versus load were plotted for the same three loads and the correspond­ing curves were drawn.

The next step was to determine the spring load that would result in the same deflections and strains as did the summer case. This was done by entering the plot (such as Figure 7) on the vertical axis with 6, £t, £VB• or £VS• A horizontal line was then drawn to intersect the appropriate curve and then drawn vertically down to the tire load axis.

The allowable spring wheel loads so determined are given in Tables 7 and 8 for the two featured test sites. For SR 2 and tires up to 12 in. wide, the re­duction in allowable load is no more than about 20 percent (from summer to spring conditions). This re­duction increases for 14- and 16-in. tires. For SR 172, the reduction in allowable load for the critical criterion (surface deflection) is about 60 percent.

A comparison of the percentage reduction between just these two pavement structures illustrates a basic difference between SR 2, which was originally de­signed to mitigate the effects of frost action, and SR 172, which was not.

Finally, Table 9 gives the allowable spring load and critical criterion for each tire size and test site. The low-volume routes such as SR 172 and SR 174 clearly have the largest reduction in allowable loads, as would be expected. On the basis of this type of analysis, actual load restrictions could be varied for each site; however, from a practical standpoint, this is not enforceable. If load re­strictions are needed for a specific pavement struc­ture, then only one or two levels of restrictions should be considered. From the analysis a spring al­lowable load of about 40 percent of the summer allow­able appears reasonable (a · 60 percent reduction). Interestingly, the corresponding allowable spring

TABLE 7 SR 2, MP 160-Spring Allowable Loads and Corresponding Percentage of the Maximum Legal Load

Maximum Spring Allow- Percent age Spring Allow- Percentage Spring Allow- Percentage Spring Allow- Percentage Legal Tire able Load for of Maximum able Load For of Maxjmum able Load for of Maximum able Load for of Maximum

Tire Size Load (lb) 0 (lb) Legal Load fr (lb) Legal Load fy B (lb) Legal Load fvs (lb) Legal Load

8-22.5 4 ,400 4,020 91 4,080 93 3,670 83 4,400 100 9-22 .5 4,950 4,600 93 4,600 93 4,190 85 4 ,920 99

10-22.5 5,500 5,050 92 5,020 9 1 4,600 84 5,390 98 I 1-22.5 6,050 5,570 92 5,830 96 4,990 82 5,900 98 12-24.5 7,920 7, 170 90 7,120 90 6,180 78 7,600 96 14-17.5 9,240 8,1 15 88 6,640 72 6,020 65 8,790 95 16-22.5 10,000 8,900 89 7,820 78 6,760 68 9,560 96

Note: 6 =su rface denection, et= horizontal s train at the bottom of the asphalt concrete, eva = vert ical strain at the top of the base, and evs =ver tical s train at the top of the subgrade,

TABLE 8 SR 172, MP 2-Spring Allowable Loads and Corresponding Percentage of the Maximum Legal Load

Maximum Legal Spring Allow- Percentage Spring Allow- Percentage Tire Load able load for o of Maximum able Load for of Maximum

Tire Size (lb) (lb) Legal Load evs (lb) Legal Load

8-22.5 4,400 1,820 41 2,330 53 9-22.5 4,950 2,180 44 2,720 55

10-22.5 5,500 2,400 44 2,980 54 11 -22,5 6,050 2,450 40 3,200 53 12-24.5 7,920 3,800 48 4,400 56 14-17,5 9,240 4,400 48 4,920 53 16-22.5 10,000 4,680 47 5,300 53

Note: 0 = surface den cc tion and evs =vertical strain at the top of the subgrnde,

66

TABLE I) Swnmary of the Critical Criteria and Corresponding Spring Allowable Load for Each Tire Size Modeled

Tire Size Site

8-22.S SR 97, MP 183.48 SR 2, MP 117.38 SR 2, MP 159.6 SR 172, MP 2.0 SR 172, MP 21.4 SR 174, MP 2,0

9-22.S SR 97, MP 183.48 SR 2, MP 117,38 SR 2, MP 159.6 SR 172, MP 2.0 SR 172, MP 21.4 SR 174, MP 2.0

10-22.S SR 97, MP 183.48 SR 2, MP 117,38 SR 2, MP 159,6 SR 172, MP 2.0 SR 172, MP 21.4 SR t 74, MP 2.0

11-22.S SR 97, MP 183.48 SR 2, MP 117.38 SR 2, MP 159.6 SR 172, MP 2.0 SR 172, MP 21.4 SR 174, MP 2.0

12-24.S SR 97, MP 183.48 SR 2, MP 117.38 SR 2, MP 159.6 SR 172, MP 2.0 SR 172, MP 21 .4 SR 174, MP 2.0

14- 17.5 SR 97, MP 183.48 SR2, MP117.38 SR 2, MP 159.6 SR 172, MP 2.0 SR 172, MP 21.4 SR 174, MP 2.0

16-22.S SR 97, MP 183.48 SR 2, MP 117.38 SR 2, MP 159.6 SR 172, MP 2.0 SR J 72, MP 21.4 SR 174, MP 2.0

Critical Cri- Spring terion for Allow able Each Site Load (lb)

0 3,775 e, s,200 €y B 3,670 0 1,820 €y B 2,400 €y!J 3,130

0 4,325 ll 5,460 eve 4,190 5 2,180 €y B 2,730 €y B 3,490

5 4,900 5 6 ,230 €y B 4,600 0 2,400 €y B 2,750 €y B 3,700

5 4,875 0 6,770 fy B 4,990 5 2,450 €y B 2,290 €y B 3,850

0 6,300 0 8~50 €y9 6,180 6 3,800 €y B 3,600 €y B 4,780

e1 6,020 5 9,380 €y B 6,020 0 4,400 €y9 3,460 €y B 4,670

f t 5,990 6 11 , 100 €y9 6,760 6 4,680 fy B 3,320 €y B 4,780

Percentage of Maxi­mum Legal Load

86 118 83 4 1 (critical) 54 7 !

87 110 85 44 (critical) 55 70

80 113 84 44 (critical) 50 67 80

112 82 40 38 (critical) 64

80 108 78 48 45 (critical) 60

65 102 65 48 37 (critical) so 60

Ill 68 47 33 (critical) 48

loads from this analysis fall within the range of the current WSDOT load restrictions (Table 1).

Criterion for When to Apply Load Restrictions

A basic issue addressed in the study was when to es­tablish load restrictions on a specific highway

30

;§: 20

~ "" >-0 "" 1i ~10

0

25 50

Transportation Research Record 1043

(assuminq that load restrictions are necessary). A criterion based on deflection measurements provides ocrtilinty ar: to the need for load rliliotrictions. 1\t

l east for the near future, it is impossible for WSDOT equipment and personnel to be at all the necessary locations during the potentially er i tical months of January, February , and March. An alternative approach i s to us e temperature data to estimate the depth of thaw in a pavement and thus whether it is near or in the critical period .

Figure 8 was prepared by calculating the depth of thaw for various thaw indices using the modified Berggren equation (]) :

X = A [(48 kavg n TI /L)Vi ]

where

x = depth of thaw (ft), A dimensionless coefficient that corrects

formula for neglected effects of volumetric

n =

heat, average thermal conductivity of the soil (Btu/ hr•ft•°F), factor for converting air thawing index to surface thawing index,

TI air thawing index (degree-days, Fahren­heit), and

L soil latent heat (Btu/ft').

The pavement structure was assumed to be homogeneous and composed of either a coarse-grained or a fine­grained soil (with corresponding dry densities of 130 and 100 pcf, respectively). An n = 1.5 was assumed (dark bituminous surface) along with A = 0. 7 for the fine-grained soil and A = 0. 6 for the coarse-grained soil . The pavement surface thickness was assumed to have a negligible effect on the depth of thaw (other than color). As shown in Figure 8 the depth of thaw for equa l thawing indices is clearly greater for coarse-gra ined soils than for fine-grained soils. Further, it is reasonable to ex­pect that the upper portions of all WSDOT pavement struct ures will behave as coarse-grained soil. Thus at an alL TI ~ 30 tlJe aepth of than will be about 12 in. and at an air TI • 50 about 15 in. For most pavement structures this would result in the surface and base courses but not necessarily all of the sub­grade being t ha wed.

The temperature data from the test sites and the BISDEF analysis of FWD data reinforce the modified Berggren calcula tions that the test sites r eached their critical condi t ion after receiving about 50

degree-days of thawing temperature. Thus it was

W=5%

h d • 130 pcl

I J 1 } yd - 100 pct

J

100 125 150 Air Thawing Index ('F-Oays)

FIGURE 8 Air thawing index versus depth of thaw for thin asphalt surfaced pavements.

Mahoney et al.

reconunended that WSOOT tentatively adopt a TI ~ 30 degree-days to indicate pavement structures ap­proaching a er i tical condition and a TI ~ 50 degree-days to indicate pavement structures in a critical condition. Clearly, pavement structure, subgrade soil, and winter temperature history will influence such criteria; however, WSDOT district maintenance personnel in the numerous maintenance offices record high and low daily temperatures for other purposes each winter. Now this same, available information can be used as a rule-of-thumb to assess the need for load restrictions.

CONCLUSIONS

The following conclusions are warranted:

1. The falling weight deflectometer is an excel­lent device for collecting the information required to evaluate the structural capacity of pavements. Further, Benkelman beam and FWD maximum deflections correlated well; however, the deflection basins ob­tainable with the FWD provide a significantly im­proved ability to analyze pavement structure.

2. For the field test sites that normally require seasonal load restrictions, the base course moduli vary more than the subgrade moduli. The subgrade moduli are relatively stable throughout the year (except when frozen). The base course weakness is

due to excessive moisture available during the thaw­ing period. The excessive moisture in the base course is exacerbated by either a still frozen subgrade or a low permeability subgrade soil (i.e., a water drainage path is temporarily reduced or eliminated), or both.

3. A multilayered elastic analysis computer pro­gram (BISDEF) was used along with FWD data to char­acterize the materials in the pavement layers for each test site with time. Criteria were developed that essentially reduce the allowable loads for a sununer condition to equivalent loads during the critical period (spring thaw) . On the basis of this analysis for the more critical test sites, a reduc­tion in legal loads of about 60 percent is required. Further, the analysis tends to reinforce the current WSOOT load restriction tables.

4. A criterion was developed that can be used to determine when load restrictions should be initiated on a pavement structure requiring such limitations

67

(the criterion does not identify which pavements re­quire load restrictions). The criterion is based on thawing degree-days and can be readily used by the WSDOT maintenance offices that record daily high and low temperatures. Both field data and an analytical procedure suggest that pavements susceptible to weakening during the critical period will approach this condition after a thawing index of 30 degree­days has occurred and will be in the critical period after accumulating 50 degree-days (one thawing degree-day is equal to an average daily temperature of 1°F above freezing). Clearly, site-specific de­flection data are the single best criterion to use in assessing the need for load restrictions, but de­flection data can be expensive to obtain and diffi­cult to get at the needed time. A temperature-based criterion is the next best alternative (and the least expensive).

REFERENCES

1. T.C. Johnson, R.L. Berg, K.L. Carey, and C.W. Kaplar. Roadway Design in Seasonal Frost Areas. NCHRP Synthesis of Highway Practice 26. TRB, Na­tional Research Council, Washington, D.C., 1974.

2. 1982 Annual Traffic Report. Washington State De­partment of Transportation, Olympia, 1982.

3. Calculation Methods for Determination of Depths of Freeze and Thaw in Soils--Emergency Construc­tion. Technical Manual 5-892-6. Department of the Army, Sept. 1966.

4. A.J, Bush III. Nondestructive Testing for Light Aircraft Pavements. Report FAA-RD-80-9-II. U.S. Army Corps of Engineers, Nov. 1980.

5. R.G. Hicks, J.D. Swait, Jr., and E.O. Chastain. Use of Layered Theory in the Design and Evaluation of Pavement Systems. 3rd ed. Department of Civil Engineering, Oregon State University, Corvallis, Jan. 1978.

6. J. Sharma, J. Hallin, and J.P. Mahoney. Evaluation of Present Legislation and Regulations on Tire Sizes, Configurations, and Load Limits. Report WA-RD-59.1. Washington State Department of Trans­portation, Olympia, July 1983.

Publication of this paper sponsored by Conuni ttee on Flexible Pavements.